Skip to content

node-tool-postgres-db

Directory actions

More options

Directory actions

More options

Latest commit

 

History

History

node-tool-postgres-db

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
src
src
 
 
.env.example
.env.example
 
 
.gitignore
.gitignore
 
 
README.md
README.md
 
 
package.json
package.json
 
 
tsconfig.json
tsconfig.json
 
 

README.md

PostgreSQL Database Tool for LLM Function Calling

This tool allows LLMs to interact with PostgreSQL databases through function calling. It supports the following operations:

  • Execute read-only SQL queries
  • Create new tables
  • Insert data
  • Update data
  • Delete data
  • Delete tables
  • List all tables
  • Get table schema information

Environment Variables Setup

Before running this tool, create a .env file and set the following environment variables:

YOMO_SFN_NAME=postgres-db-tool YOMO_SFN_ZIPPER=zipper.vivgrid.com:9000 YOMO_SFN_CREDENTIAL=<your-yomo-sfn-credential> # PostgreSQL database connection URL DATABASE_URL=postgres://username:password@localhost:5432/database_name

Installation

npm install

Running

1. Install YoMo CLI

curl -fsSL https://get.yomo.run | sh

Detailed CLI usage instructions can be found in the YoMo CLI documentation.

2. Run the function

yomo run -n postgres-db-tool

Function Call Examples

Here are some examples of how to use this tool:

curl --request POST \ --url https://api.vivgrid.com/v1/chat/completions \ --header 'Authorization: Bearer <YOUR_TOKEN>' \ --header 'content-type: application/json' \ --data '{  "messages": [  {  "role": "assistant",  "content": "tell me how many tables in my centre database"  }  ] }'

You will get following result:

{ "id": "chatcmpl-B9M9aiAAv0Q3rf6LxxUFxphuAVPsC", "object": "chat.completion", "created": 1741569810, "model": "gpt-4o-2024-08-06", "choices": [ { "index": 0, "message": { "role": "assistant", "content": "Your \"centre\" database contains 5 tables." }, "finish_reason": "stop", "content_filter_results": { "hate": { "filtered": false, "severity": "safe" }, "self_harm": { "filtered": false, "severity": "safe" }, "sexual": { "filtered": false, "severity": "safe" }, "violence": { "filtered": false, "severity": "safe" }, "jailbreak": { "filtered": false, "detected": false }, "profanity": { "filtered": false, "detected": false } } } ], "usage": { "prompt_tokens": 164, "completion_tokens": 45, "total_tokens": 117, "prompt_tokens_details": { "audio_tokens": 0, "cached_tokens": 0 }, "completion_tokens_details": { "audio_tokens": 0, "reasoning_tokens": 0 } }, "system_fingerprint": "fp_b705f0c291", "prompt_filter_results": [ { "index": 0, "content_filter_results": { "hate": { "filtered": false }, "self_harm": { "filtered": false }, "sexual": { "filtered": false }, "violence": { "filtered": false }, "jailbreak": { "filtered": false, "detected": false }, "profanity": { "filtered": false, "detected": false } } } ] }

Security Considerations

  • Ensure the database user has only necessary permissions
  • Consider using a read-only user for query operations
  • In production environments, don't hardcode database credentials directly in the code
  • Consider using connection pool limits and timeout settings to prevent resource exhaustion